Aphasia is among the most common and disabling consequences of stroke. Language rehabilitation is frequently effective, but recovery is often incomplete and slow. A few small studies have shown that medications, including cholinesterase inhibitors and selective serotonin reuptake inhibtors (SSRIs), can augment language rehabilitation to enhance aphasia recovery. These investigations have been based on strong theoretical grounds, as cholinesterase inhibitors and SSRIs increase the availability of neurotransmitters that are essential for neuroplasticity. But results have been inconsistent, with small effects. Aphasia treatment could be substantially advanced by a large randomized trial showing clinically significant benefit of specific medications, along with language intervention. However, clinical trials are premature because the most effective medications and the characteristics of individuals most likely to respond have not yet been identified. In previous funding cycles, we have shown that the volume and location of stroke, other medications that might block the effect of the tested medication, and the health of the uninfarcted tissue might strongly affect the response to treatment. Furthermore, it is essential to base the timing of intervention on empirical evidence regarding the time at which intervention has the greatest effect. We propose to obtain all of this essential information to design a rational randomized trial of medications (versus placebo), along with language therapy, that will have the greatest probability of identifying clinically important and reliable effects. We will obtain this new information in a prospective, longitudinal study of language recovery and multimodality brain imaging at four time points after stroke (Week 1, Month 3, Month 6, and Month 12). We will not attempt to manipulate medications, as there are insufficient data on their effects. Rather, medication use (prescribed at the discretion of the physician team) will be recorded at each visit, and confirmed through pharmacy records and pill counting. Our progress and record of enrolling large numbers of participants show that we will have sufficient power to find any significant effects. The structural and resting state fMRI studies will help us understand the mechanisms of recovery (or decline) after stroke. We hypothesize that beneficial medications will enhance connectivity between critical nodes in the language network, while medications with a negative effect (by blocking neurotransmitters or their receptors) will be associated with decrease in connectivity between these nodes. The brain imaging and behavioral data together will also help us achieve another important goal, aligned with the mission of NIDCD, to improve prognosis for aphasia recovery at an individual level. Identifying lesion and brain characteristics, medications, demographics, and number of rehabilitation sessions that independently contribute to aphasia recovery will allow us to better predict who is likely to achieve the greatest recovery and under what conditions, as well as provide the critical basis for a well- designed, multicenter clinical trial that will likely have an important impact on aphasia rehabilitation.